Hamidou Dia

Just wrapped our earnings call and wanted to share a few highlights from the quarter. Bottom line: We’ve got great momentum across our AI platform and our family of Copilots, fueling our increasing investments in both capital and talent. We’re building a planet-scale cloud and AI factory. We’ll increase our AI capacity by 80% this year, and nearly double our DC footprint over the next two. That includes Fairwater in Wisconsin, the world’s most powerful AI datacenter, which will scale to two gigawatts alone. We are building a fungible fleet that spans every stage of the AI lifecycle and is constantly optimized to deliver the best ROI and TCO for us and our customers. One example: during the quarter, we increased the token throughput for GPT-4.1 and GPT-5 by more than 30% per GPU. Our family of Copilots and agents also surpassed over 150 million monthly active users across high value domains: • Information work: 90 percent of the F500 now use Microsoft 365 Copilot – and we’re shipping updates very fast • Coding: GitHub Copilot now has 26 million-plus users • Security: 36+ agents integrated across Entra, Defender, Purview, Intune • Science & Health: We helped document 17 million-plus patient encounters, up 5X year over year • Consumer: Tons of new capabilities announced over the past two weeks, from Groups and Mico in Copilot, to "Hey Copilot" in Windows And this week, we signed a new agreement with OpenAI, a milestone for both companies. It’s been an incredible investment, with nearly 10X return and no carry. All the value goes straight to shareholders. Another thing I feel super proud about is that Microsoft is now behind two of the largest nonprofits in the world! You can read more about our results here: https://lnkd.in/gxtVeMk2

7,824

443 Comments

Amazon Web Services (AWS) today announced a series of innovations that make it easier, faster, and more accessible for adopting and scaling AI and agentic systems. Of the many launches our CEO Matt Garman presented here at re:Invent 2025, I'd highlight three: → #FrontierAgents, a new class of fully autonomous, highly scalable AI agents that can work for hours or days without intervention. #Kiro serves as the virtual developer, AWS Security Agent acts as a security consultant, and AWS DevOps Agent operates as an on-call operations team. → #NovaForge, a solution that empowers companies of any size to build their own AI models by combining proprietary data with Amazon Nova-curated datasets. → #AmazonEC2Trainium3UltraServers, a new generation of processors delivering up to 4.4x more computational performance, 4x greater energy efficiency, and nearly 4x more memory bandwidth. And here's the kicker: we are already working on Trainium4. Today's announcements also include the largest expansion of new models to date in Amazon Bedrock's history, with the arrival of Google’s Gemma 3, MiniMax’s M2, NVIDIA’s Nemotron, OpenAI’s GPT OSS Safeguard, etc. Plus, we launched new Bedrock AgentCore capabilities, Nova 2 models, AWS AI Factories, AWS Security Hub, Database Savings Plans—a flexible pricing model that cuts costs by up to 35%—and more. These aren't just incremental improvements; they represent a phase shift in how we develop technology. It's the beginning of a new innovation cycle that will further accelerate business transformation and democratize access to cutting-edge AI for companies of all sizes around the world and, of course, in Latin America! Check out all the details here: https://lnkd.in/dbtF4EYP #AWS #AWSLATAM #reInvent #reInvent2025 #Cloud #AI #GenAI

317

1 Comment

Strong look at how our differentiated portfolio and customer partnerships are driving Google Cloud's growth story coming out of the Goldman Sachs' Communacopia conference yesterday: https://lnkd.in/gktdMFy4 As Thomas says, "We've spent years building advanced AI technology of our own — chips, systems, tools, agents — we made those bets very early. Much of the work that you see today has been underway for many, many years. We're not just reselling third-party AI technology. So why we're winning is because we see this deep product differentiation now being adopted by customers and (broadening) our addressable market. In turn, that's leading us to grow revenue and operating income."

242

3 Comments

IBM CEO Arvind Krishna shatters the AI replacement myth: 30% of code will be AI-generated, but we'll need MORE developers, not fewer. This week at AI ALPI, we conducted a deep-dive analysis of Arvind Krishna's insights that he shared at SXSW that HR leaders must understand as AI reshapes workforce planning: → The productivity paradox: When developers become 30% more productive through AI, history shows market leaders capture expanded market share, driving increased demand for talent—not reduction ↳ Organizations approaching AI as a replacement tool rather than an amplification engine are leaving significant competitive advantage on the table → Enterprise data arbitrage: 99% of proprietary enterprise data remains unlocked from AI models ↳L'Oréal's 150 years of cosmetic science data represents the pattern for industry leaders—companies that integrate domain-specific AI with proprietary data will create insurmountable competitive moats → The coming talent war: Quantum computing advances will drive transformational breakthroughs before 2028 in materials science, molecular modeling, and optimization algorithms ↳ HR leaders who begin developing quantum literacy pathways now will have 3-5 year advantages in talent acquisition as these skills become critical infrastructure → AI efficiency revolution: Current AI models will achieve 99% energy reduction within 5 years ↳ This will democratize AI access but dramatically increase competition for specialized talent who can architect these systems → Global talent acquisition: Successful innovation ecosystems require bringing together academic institutions, government funding, and specialized talent—with Illinois winning 4 of 10 DOE quantum grants by making strategic investments ↳ Immigration restrictions may significantly hamper US competitiveness in emerging fields where talent flows to the centers of innovation While everyone focuses on layoffs during AI implementations, Krishna revealed the counterintuitive pattern: after initial productivity gains, companies that successfully implement AI technology ultimately expand their tech workforces by 40-60% within 24-36 months. This pattern was visible in previous technological revolutions from calculators to Photoshop, where "labor-saving" technologies created entirely new markets. 🔥 Want more breakdowns like this? Follow along for insights on: → Getting started with AI in HR teams → Scaling AI adoption across HR functions → Building AI competency in HR departments → Taking HR AI platforms to enterprise market → Developing HR AI products that solve real problems

226

204 Comments

Our commitment at Google Cloud is to bring our leading AI capabilities to customers where they are. I'm thrilled to announce our expanded partnership with Oracle, which will make our #Gemini models available to customers via #VertexAI and the Oracle Cloud Infrastructure Generative AI service. Oracle is also collaborating with Google Cloud to make Gemini models available within Oracle Fusion Cloud Applications. This collaboration is a significant win for enterprises, enabling them to bring Google Cloud's powerful AI directly into their core data platforms and business workflows. https://lnkd.in/gPdtPXCH

83

People often ask me about the differences in architecture between Amazon Dynamo (the 2007 SOSP paper), DynamoDB (the AWS serverless NoSQL database), and Aurora DSQL (the AWS serverless SQL databases). I captured the key points on durability, high availability, programming model, and trade-offs in my blog: https://lnkd.in/gkMG_tVD TL;DR: Despite the shared name, Dynamo and DynamoDB have little in common architecturally, and make very different trade-off decisions.

632

8 Comments

Agentic AI will create a ~$1 trillion global market for agentic AI services globally. This is an incredible opportunity for our partner ecosystem to lead and define the category. 90%+ of enterprises are planning to deploy agentic AI in the next three years. To seize this opportunity, software and services firms must evolve. Our latest analysis includes a strategic framework to shift from implementation to co-designing transformative workflows and innovating commercially. Read my thoughts on where the biggest opportunities are across different industries to develop custom agents on our leading AI platform, commercialize through the Google Cloud Agent Marketplace and make available to users on Agentspace: https://lnkd.in/gjF_PZxc

425

29 Comments

In our CTO organization, we work with hundreds of customers and across our Google Cloud product, engineering and go to market teams on the horizons of where tech will help solve previously intractable challenges. One of the hardest things for our customers, partners and even internal teams to reason about is how the vast set of product launches might be useful to them. So I asked my team to share a few perspectives in our latest #AskOCTO column related to Gemini 2.0, and how all the capabilities under the hood (infra, platform, models, agents) might benefit our customers. Hope you'll find some of them useful as conversation/action starters in your own org. And since building is always the best way to learn, definitely recommend you swing by aistudio.google.com to try Gemini 2.0 for yourself. https://lnkd.in/gBVnG3pQ thx to Scott Penberthy, Ben McCormack, Antonio Gulli, Jeff Sternberg, Olaf Schnapauff, Ashwin Ram, Jack Ngare, Lee Boonstra, Dare Kolawole, Julian Sniffen and Troy Trimble for sharing your experiences.

160

3 Comments